SkyRL Boosts Qwen3.5-397B Agent Pass@1 from 16% to 27% via RL

Mercor Research and the SkyRL team detail a DPPO-based RL post-training recipe that lifts Qwen3.5-397B's Pass@1 on APEX-Agents long-horizon office tasks from 16% to about 27%.

2026-09-02 ~ 2026-09-03 · 3 related posts

• RL Training Guide: Qwen 3.5 397B Pass@1 Boosted to 27.3% — mariofilhoml · 2026-09-02
• SkyRL Trains Qwen3.5-397B on Long-Horizon Agent Tasks, Pass@1 Jumps from 16% to 27% — simonguozirui · 2026-09-03
• SkyRL field guide: RL training lifts Qwen3.5 agent Pass@1 from 16% to 27% — SinclairWang1 · 2026-09-03